Data Management Protocolv1.0 · April 2026

Evidence you can
trace to its source.

A six-section protocol that turns field evidence into traceable findings - anchored in context analysis, executed through seven complementary data-collection instruments, and held together by an encrypted platform that links every report citation back to its source.

Operating Principles

How we hold the chain of evidence together

P.01

Context first

Every dataset is shaped by where it was collected. Sampling validity depends on contextual understanding - conflict, displacement and exclusion factors all introduce bias that statistics alone cannot correct.

P.02

Traceable evidence

Every finding links back to its source - a KII transcript, a georeferenced photo, a dashboard filter, a household submission. Reports become living interfaces, not closed documents.

P.03

Triangulated confidence

Findings are scored on the frequency of corroborating reports across instruments and respondent profiles. Photos and videos stand on their own; other claims require independent confirmation.

P.04

Quality assurance along the chain

Review of data collection and analysis is applied at every step - KoBo coherence checks, supervisor sampling verification, duration outliers, and same-day deviation flags.

01

Section

Context first.

Every dataset is shaped by where it was collected. Before findings, the operating environment.

Axiom ME treats context analysis as a precondition for credible findings - not an optional preamble. Sampling validity depends on contextual understanding. Conflict, displacement and exclusion factors all introduce bias that statistical methods alone cannot correct for.

Demographic & clan mapping across the area of operation

Unresolved vs. resolved conflicts and their spatial expression

Religious, ethnic and political fault lines affecting respondents

Informed sampling and access strategy tuned to security

02

Section

Seven instruments,
one evidence base.

Each method has a specific evidentiary purpose. Together they triangulate findings across quantitative, qualitative, observational and discreet dimensions.

7
Field instruments
50K+
Household interviews
20K+
Geo-referenced photos
5
Countries covered
2.1
Qualitative

Key Informant Interview

- KII

Semi-structured conversations with purposively selected stakeholders, designed to surface verified, context-rich evidence.

2.2
Quantitative

Face-to-Face Household Survey

- HH

Random-sample interviews delivered in person, supported by KoBo on tablets and rigorous statistical design.

2.3
Qualitative

Focus Group Discussion

- FGD

Collective conversation, separated by gender and age, in a safe and culturally appropriate setting.

2.4
Qualitative

Case Studies

- Case

Human-interest stories from remarkable beneficiaries - surfacing what aggregate data cannot show.

2.5
Quantitative · Remote

Computer Assisted Telephone Interview

- CATI

Dedicated call-centre facilities in Hargeisa and Sanaa, scalable to fifty specialised enumerators.

2.6
Observational

Field Observations

- FO

On-the-ground verification of implementation against protocols, supported by photo and video evidence.

2.7
Sensitive

Ghost Monitoring

- Ghost

Discreet research with embedded local researchers, surfacing diversion or fraud without putting enumerators at risk.

2.1

Key Informant Interview

- KII

Conditions of engagement - use of information, confidentiality, quality of feedback, the right to decline - are outlined to secure informed consent. Upon confirmation, georeferenced photos of the interview are taken. A standard set of questions across stakeholder types eases triangulation and surfaces variation in leading answers.

2.2

Face-to-Face Household Survey

- HH

Bio-information of head of household, number of children, age and gender of interviewed children, finder marks, refusal reasons, displacement patterns and recipient status are all collected. The KoBo system confirms location and timing - supervisors can verify sampling adherence and detect duplicates in real time. Over 50,000 household interviews delivered across Somalia, Ethiopia, Kenya, South Sudan and Yemen.

2.3

Focus Group Discussion

- FGD

Separate FGDs are conducted for women, youth and men. Two enumerators per session - one facilitating, one taking notes. Bio-information on each member is collected for QA. The profile of respondents - IDPs, returnees, refugees, migrants, host communities, minority members, rural versus urban - is selected based on the targeted population.

2.4

Case Studies

- Case

Case studies are drawn from highly vulnerable, exceptionally successful, or otherwise singular beneficiaries. Continuous consent, anonymity by default, silhouette images when needed. Each story is reviewed for potential consequences before publication.

2.5

Computer Assisted Telephone Interview

- CATI

The call centre handles security-restricted or inaccessible areas. Enumerators are selected by skills, gender, dialect and location. Sampling protocols define the list of beneficiaries to be called, the use of replacement lists, and response-rate scoring. SIM-swap awareness shapes call timing; live data entry with real-time coherence checks; interviews above 45 minutes flagged for review.

2.6

Field Observations

- FO

Inspection of beneficiary registers, distribution logs and incident reports. Direct observation of ID verification, distribution protocols, hygiene measures, security posture. Each photo is timestamped, georeferenced (when safe to do so) and linked to a structured questionnaire. Over 20,000 georeferenced field photographs archived.

2.7

Ghost Monitoring

- Ghost

Up to three independent researchers may be assigned to the same question - none informed of the others - and their reports are compared to clarify event versus pattern. The approach takes the form of informal conversation rather than structured interview. No names, no phone numbers. Reports remain fully confidential and outside the Axiom ME public track record.

03

Section

The platform behind the findings.

Technology, data management, and the systems that handle photos, videos and documents - turning field evidence into traceable findings.

Technology support & hardware

KoBo on tablets/phones for face-to-face and FO tools, with mandatory fields and conditional logic enforced at entry. CATI workstations with full-time supervisors and call recording. High-quality cameras for evidentiary photography.

Data management, processing & analysis

Web-based platform organising documents per project and per phase - desk review, field data, flash reports, minutes - with upload metadata, version tracking and direct-link access from the report.

Encrypted document & photo platform

Encrypted website hosts desk review, KII transcripts, FGD notes, case studies, IP tools, field reports, FO checklists, and the photo/video archive. Each link is referenced from the report itself, within the limits of confidentiality and respondent consent.

Photos & videos as evidence

Each photo lands on satellite imagery at its captured location, with timestamp and GPS metadata, supporting verification of the field visit. The system handles very large numbers of photos without saturating report size.

04

Section

Creative rigour.
Traceable evidence.

Data analysis is a creative and intuitive task. Rigorous methodology and lateral thinking surface findings that a single-axis read would miss.

Software stack

MS Excel · LibreOffice
Statistical analysis, standalone interactive dashboards, complex analysis.
Python
AI pattern detection across multiple databases; Twitter trend analysis (e.g. OCHA Somalia Anticipatory Survey, 2021).
R
Online dashboards and statistical analysis (IOM Somalia Baseline).
STATA
Statistical data analysis.
NVivo
Text and qualitative analysis.
Pajek · Graphviz
Social Network Analysis - snowball-sampled stakeholder mapping, cluster and gatekeeper identification.
QGIS · ArcGIS
Spatial analysis of georeferenced household and infrastructure data.
Power BI
UN agency dashboards in Somalia with monthly datasets, filters and map visualisation of data-collection locations.
KoBo · ODK
Tablet-based field data collection with mandatory fields, conditional logic and real-time coherence checks.
05

Section

One encrypted home for every document.

Excel, Word, PDF, photo and video - reorganised and accessible via URL link from the encrypted Axiom ME website. Reports become “light” deliverables that link out to every supporting document.

Index of Confidence in Findings

Level of confidence in reported information is defined using the frequency of similar reports across the different tools and per category of respondent within the same tool.

Triangulation

Number of independent sources that confirm the same information.

Accuracy

Statistical power of the sample; focus on qualitative tools; level of detail in each response.

Repetition

How often the same information surfaces across instruments and respondent profiles.

Photos as virtual field visits

Each photo lands on satellite imagery at the location captured. Readers can zoom into the surrounding area and inspect the operational context - roads, infrastructure, settlements.

Reporting against the Logical Framework

Data is organised as interactive dashboards measuring LF indicators automatically. Each indicator is recalculated in an evidence-based manner and benchmarked against IP-reported values.

06

Section

The dashboard
is the companion.

Every Axiom ME field report follows a consistent structure - context, methodology, findings, lessons learnt, recommendations - with the analytical dashboard as a live companion.

“Data is evidence; analysis is interpretation; reporting is translation. The protocol is what holds them together and keeps every finding traceable.”

- Axiom ME · Data Management Protocol · Closing note
  1. 6.1

    Context analysis

    The operating environment and how it shaped data collection.

  2. 6.2

    Profiling

    Value-chain actors based on the beneficiary list provided by the client.

  3. 6.3

    Methodology

    Tools, sampling, limitations and validity.

  4. 6.4

    Key findings

    Headline observations and the evidence behind each.

  5. 6.5

    Analysis as dashboards

    All questions analysed, with concatenated views for cross-cutting reads.

  6. 6.6

    Lessons learnt

    What works very well, and what does not.

  7. 6.7

    Recommendations

    Anchored in lessons learnt, written for the client's decision-making moments.

From protocol to delivery

How the protocol shows up across our services

The Data Management Protocol is the common backbone of every Axiom ME mandate - from large-scale Third-Party Monitoring to evaluations, context analyses, and capacity-building engagements.

Explore Services

Work With Us

Working in fragile states?
Let's build the evidence
together.

Whether you need independent field verification, rigorous evaluation, or deep context analysis - we have the network, the methodology, and the field presence to deliver.

Submit a ToR

Every inquiry acknowledged within 48 hours